Amidine-Mediated Zwitterionic Ring-Opening ... - ACS Publications

Feb 5, 2016 - LiLin He,. ‡. Changwoo Do,. ‡. Jayne C. Garno,. † and Donghui Zhang*,†. †. Department of Chemistry and Macromolecular Studies ...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/Macromolecules

Amidine-Mediated Zwitterionic Ring-Opening Polymerization of N‑Alkyl N‑Carboxyanhydride: Mechanism, Kinetics, and Architecture Elucidation Ang Li,† Lu Lu,† Xin Li,† LiLin He,‡ Changwoo Do,‡ Jayne C. Garno,† and Donghui Zhang*,† †

Department of Chemistry and Macromolecular Studies Group, Louisiana State University, Baton Rouge, Louisiana 70803, United States ‡ Biology and Soft Matter Division, Neutron Sciences Directorate, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, United States S Supporting Information *

ABSTRACT: Zwitterionic ring-opening polymerization (ZROP) of Nbutyl N-carboxyanhydrides (Bu-NCAs) has been investigated using 1,8diazabicycloundec-7-ene (DBU), a bicyclic amidine initiator. It was found that poly(N-butylglycine)s (PNBGs) with molecular weight (Mn) in the 3.5−32.4 kg mol−1 range and polydispersity index (PDI) in the 1.02−1.12 range can be readily obtained by systematically varying the initial monomer to initiator feed ratio. The polymerization exhibits characteristics of a controlled polymerization, as evidenced by the linear increase of polymer molecular weight with conversion and the successful enchainment experiments. Kinetic studies revealed that the reaction is first-order dependent on the monomer and the DBU concentration. The rate of initiation is comparable to that of the propagation. Random copolypeptoids of poly[(N-propargylglycine)-r-(N-butylglycine)]s [P(NPgG-r-NBG)s] were also synthesized by DBU-mediated copolymerization of Bu-NCA and N-propargyl N-carboxyanhydride (Pg-NCA). Subsequent grafting with azido-terminated poly(ethylene glycol) (PEG) produces bottlebrush copolymers. Analysis of bottlebrush copolymer samples using atomic force microscopy (AFM) revealed a surface morphology of toroid-shaped nanostructures, consistent with the polypeptoid backbone having cyclic architecture. Small-angle neutron scattering (SANS) characterization of the bottlebrush polymer ensemble in solution also confirms the cyclic architecture of the polypeptoid backbones.



transition windows22 relative to the linear counterparts. To date, several synthetic approaches have been developed for the cyclic polymers, including interfacial condensation,23 ringexpansion metathesis polymerization,19,24 end-to-end polymer cyclization,25,26 and zwitterionic ring-opening polymerization (ZROP).27−30 The ZROP method entails the formation of a zwitterionic propagating intermediate where the two chain ends are oppositely charged and interact strongly with one another by electrostatic interaction.31 The zwitterionic intermediate can react either with monomers (by a chain growth mechanism) or with each other (by a step growth mechanism) to elongate the chain or undergo intramolecular chain transfer to yield a mixture of linear and cyclic polymers (via backbiting mechanism). Intramolecular chain transfer that occurs exclusively in an end-to-end fashion will produce macrocyclic polymers cleanly. While the end-to-end macrocyclization is

INTRODUCTION Polypeptoids, also known as poly(N-substituted glycine)s, are synthetic mimics of polypeptides. Because of the Nsubstitution, polypeptoids lack main chain chirality and hydrogen bonding interaction, in contrast to polypeptides. This results in polypeptoid having enhanced proteolytic stability relative to polypeptides,1 thermoprocessability,2−4 and conformational tunability (via side chain structure).5−8 Secondary conformations of polypeptoids including random coil, helices, and sheet structures have all been reported.9−13 In addition, polypeptoids have been shown to exhibit minimal cytotoxicity to several cell lines14−16 and can degrade under oxidative conditions that mimic the tissue inflammation environment, suggesting the potential uses for in vivo biomedical applications.17 Cyclic polymers have attracted considerable attention because of their distinctive physical properties as compared to the linear counterparts. The unique topology imparts many interesting properties to cyclic polymers such as lower intrinsic viscosity,18,19 higher glass transition and melting temperature,4,20 faster crystallization rates,21 and wider LCST phase © XXXX American Chemical Society

Received: December 1, 2015 Revised: January 23, 2016

A

DOI: 10.1021/acs.macromol.5b02611 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Scheme 1

Table 1. DBU-Mediated Polymerization of Bu-NCA in THF with Different Initial Monomer-to-Initiator Ratios ([M]0:[DBU]0)a entry no.

[M]0:[DBU]0

[M]0b (M)

Mn(theor)c (kg mol−1)

timea (h)

Mn(SEC)d (kg mol−1)

PDId

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

25:1 50:1 75:1 100:1 150:1 200:1 400:1 25:1 50:1 100:1 200:1 400:1 50:1 100:1 200:1 400:1

0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.8 0.8 0.8 0.8 0.8 1.6 1.6 1.6 1.6

2.9 5.8 8.6 11.4 17.1 22.7 45.4 2.9 5.7 11.4 22.7 45.4 5.7 11.4 22.7 45.4

4 8 12 16 24 32 64 2 4 8 16 32 1 2 4 8

3.5 5.9 8.4 9.7 14.1 18.4 32.4 3.8 5.7 9.5 16.6 19.3 4.6 8.5 15.2 26.5

1.08 1.10 1.02 1.05 1.08 1.08 1.04 1.12 1.10 1.06 1.02 1.11 1.07 1.04 1.02 1.03

All reactions were conducted in 50 °C THF for different duration (1−64 h) to reach complete conversions. bThe initial monomer concentration. The theoretical polymer molecular weight is calculated using the initial monomer to DBU ratio ([M]0:[I]0) assuming a single-site initiation by DBU and complete conversions. dExperimental polymer molecular weights were obtained by SEC-MALS-DRI method (DMF/0.1 M LiBr, 25 °C) with a measured dn/dc = 0.0797(9) mL g−1. a c

mers.3,5,16,36 The reactions were shown to occur in a controlled manner, producing cyclic polypeptoids with well-defined structure, molecular weight, and narrow molecular weight distribution. This is in contrast to early studies on pyridine/ tertiary amine/solvent initiated or thermally initiated ZROPs of Me-NCA which only produce low molecular weight cyclic poly(N-methyl glycine) (a.k.a. polysarcosin).37−40 1,8-Diazabicycloundec-7-ene (DBU), a bicyclic amidine, is widely used to catalyze esterification of carboxylic acids41 and controlled ROPs of lactides, lactones, and cyclic phosphoesters.42−45 The DBU is commonly considered a strong base with weak nucleophilicity. This view has changed in recent years after DBU was found to be a competent nucleophile that can undergo nucleophilic addition to dimethyl carbonate and chloroformate.46−49 It was recently demonstrated that DBU can initiate the ZROP of lactide to produce a mixture of cyclic and linear polylactide.50 While NHCs can mediate the ZROP of various cyclic substrates, they are air and moisture sensitive. In an effort to identify and characterize alternative organic initiators that can outperform NHCs in mediating the ZROP of R-NCAs to produce cyclic polypeptoids in a controlled manner and ideally more robust than NHCs, we investigated the DBU as a possible initiator for the ZROP of Bu-NCA monomers. It was found that the polymerization occurs efficiently in a controlled fashion, producing cyclic poly(N-butylglycine)s (PNBGs) with good molecular weight control and narrow molecular weight distribution. The polymerization rate is slightly faster to those of previously reported NHC-mediated ZROP of Bu-NCA. In views of the enhanced moisture/air stability and availability of

entropically unfavored, high molecular weight cyclic polymers can be cleanly produced by the ZROP method when the interactions between the oppositely charged chain ends of the zwitterionic propagating intermediates overcome the entropic penalties.31 A distinct advantage of the ZROP method is that the polymerization can be conducted in moderate to high monomer concentration, enabling efficient access to cyclic polymers. This is in contrast to the end-to-end cyclization strategy of linear polymer precursors that typically require high dilution conditions. ZROP has recently emerged as a promising strategy to produce high molecular weight cyclic polymers with diverse backbone structures (e.g., polyester27,28,30,32 and polyamide).5 For example, it was recently reported that isothiourea can mediate the ZROP of lactide to produce cyclic polylactides with relatively high cyclic polymer contents in spite of somewhat slow polymerization rates.33 B(C6F5)3 was used as the electrophilic initiator to mediate the ZROP of glycidyl monomers to yield cyclic polyethers.34 Several recent studies have shown that N-heterocyclic carbenes (NHCs) can initiate and mediate the ZROPs of strained cyclic monomers (e.g., lactides, caprolactone, and carbosiloxane) to produce high molecular weight macrocyclic polylactides, polylactones, and polycarbosiloxanes with moderate to high efficiency.27−30 Cyclic gradient polyesters can be obtained by NHC-mediated ZROP of ε-caprolactone (CL) and δ-valerolactone (VL) in a one-pot reaction.35 We have recently demonstrated that a variety of cyclic poly(N-substituted glycine) (a.k.a. polypeptoids) with varying N-substituents can be obtained by NHCmediated ZROP of the corresponding R-NCA monoB

DOI: 10.1021/acs.macromol.5b02611 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 1. (A) Representative normalized SEC chromatograms of PNBG obtained at varying initial monomer to initiator ratios (entries 1−7, Table 1). (B) Plots of experimental Mn (■) and PDI (blue ●) versus conversion of DBU-mediated polymerization of Bu-NCA ([M]0:[I]0 = 100:1, [M]0 = 0.4 M, 50 °C, THF), the linear fit of the Mn versus conversion plot (red ) (R2 = 0.98), and theoretical trend line of a living polymerization (···).

ratios ([M]0= 0.4 M, entry 1−7, Table 1). The polymer molecular weight distribution remains relatively narrow (PDI = 1.02−1.12) (Figure 1A). The experimental M n values determined by the SEC analysis agree reasonably well with the theoretical values based on single site initiation by DBU when the initial monomer to initiator ratio ([M]0:[DBU]0) is lower than 150:1. However, when the [M]0:[DBU] ratio exceeds 150:1, the experimental Mn values become less than theoretical values, suggesting side reactions in competition with chain propagation. If the DBU initiates the polymerization by a nucleophilic ring-opening pathway, it is conceivable that intramolecular macrocyclization or intermolecular head-to-tail chain coupling may occur, resulting in the formation of free DBUs. The liberated DBU can reinitiate the chain growth, resulting in reduced polymer molecular weights. Alternatively, residual nucleophilic impurity such as H2O can also initiate the polymerization, giving rise to reduced polymer Mns than theoretical values, consistent with the MALDI-TOF MS results suggesting the presence of PNBGs composed of one carboxylic end-group and one secondary amino end-group in low apparent content. Controlled experiments showed that while H2O can initiate the polymerization of Bu-NCA under the standard polymerization conditions (i.e., [M]0:[H2O]0 = 50:1, THF, 50 °C), the initiation is significantly slower than that by other nucleophilic initiators (e.g., primary amine or DBU). This is supported by the much longer reaction time to reach complete monomer consumption and the significantly higher polymer molecular weight of the resulting PNBG when water is the initiator than when primary amine or DBU are used as initiators. It was also noted that polymerization conducted at higher initial monomer concentration ([M]0 = 0.8 or 1.6 M) and initial monomer to initiator ratio ([M]0:[DBU]0 = 400:1) (entries 12 and 16, Table 1) yielded PNBG polymers whose Mn deviates more significantly from the theoretical value than that from the lower initial monomer concentration (entry 7, Table 1). The origin for the deviation is presently unclear. The polymer molecular weight was found to increase linearly with conversion during the DBU-mediated polymerization of Bu-NCA in THF (Figure 1B). The polymer molecular weight distribution also remains narrow PDI (PDI = 1.02−1.03) throughout the polymerization. However, the Mn versus conversion plot does not go through the (0,0) origin, suggesting that the rate of initiation is either comparable to or slower than that of the propagation in the polymerization. 1 H NMR analyses of the polymerization reaction mixtures upon

DBU relative to the NHCs, the reported polymerization method represents an attractive synthetic approach toward cyclic polypeptoids. AFM and SANS analysis of cyclic bottlebrush polymers consisting of polypeptoid backbone and poly(ethylene glycol) (PEG) side chains have been conducted to verify the cyclic architecture of polypeptoids produced by the DBU-mediated ZROP method.



RESULTS AND DISCUSSION DBU-Mediated Polymerization of Bu-NCAs. A series of polymerizations of Bu-NCA in the presence of DBU were conducted in 50 °C THF with varying initial monomer to DBU feed ratios ([M]0:[DBU]0 = 25:1 to 400:1) at three different initial monomer concentrations ([M]0 = 0.4, 0.8, and 1.6 M). Progression of the polymerization was monitored by 1H NMR spectroscopic analysis of the reaction aliquots taken at different time. All reactions were allowed to reach a complete conversion prior to termination by precipitation of the polymers in hexane. The isolated polymers were characterized by 1H NMR, MALDI-TOF MS spectroscopy, and size-exclusion chromatography coupled to multiangle light scattering and differential refractive index detectors (SEC-MALS-DRI). 1H NMR analysis revealed the desired PNBG backbone structure and the presence of DBU moieties affixed to the polymer chains, as evidenced by the presence of two methylene groups (f, g) of the DBU moieties at 2.84 and 2.05 ppm, respectively (Figure S1). MALDI-TOF MS analysis of a low molecular weight polymer revealed the dominant presence of molecular ions that are consistent with the structures of cyclic PNBG polymers bearing one DBU moiety (a, b, c, and h, Figure S2) and the cyclic zwitterionic PNBG species bearing one DBU moiety and one carbamate end-group (d and e, Figure S2) (Scheme 1). Minor sets of molecular ions that are consistent with the linear PNBG polymers bearing one carboxylic acid end-group and one secondary amino end-group were also present (f and g, Figure S2). In addition, some of the cyclic PNBG polymers were co-ionized with methanol (the solvent introduced during the MALDI-TOF MS sample preparation) under the MALDITOF MS conditions (Figure S2). Similar behaviors have previously been reported for the MS studies of the PNBG polymers obtained from the NHC-mediated ZROP of BuNCAs.51 SEC-MALS-DRI analysis revealed that PNBGs with predicable Mn values in the 3.5−32.4 kg mol−1 range can be readily synthesized by controlling the initial monomer to DBU C

DOI: 10.1021/acs.macromol.5b02611 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 2. (A) Plots of ln([M]0:[M]) versus time for DBU-mediated polymerization of Bu-NCA in toluene-d8 at 50 °C and the linear fitting of the data (gray line). (B) plot of observed polymerization rate constants (kobs) versus the initial DBU concentration and the linear fitting of the data (red line) (R2 = 0.998). (C) Plot of concentration of polymerized Bu-NCA ([M]p) versus reaction time of DBU-mediated polymerization within the first 25% conversion ([M]0 = 0.15 M and 50 °C in toluene-d8). (D) Plot of conversion of DBU-mediated polymerization of Bu-NCA versus reaction time of DBU-mediated polymerization of Bu-NCA up to 80% conversion. The red lines in (C) and (D) are the fitting of the data (C: R2 = 0.997; D: R2 = 0.983) using a kinetic model describing a living polymerization with slow initiation relative to propagation.

to initiator ratios ([M]0:[DBU]0 = 25:1−150:1, Figure 2A). The plot of ln([M]0:[M]) versus time for all reactions can be linearly fitted quite well, indicating that the polymerization is first-order dependent on the monomer concentration. The slope of the linear fits afforded the observed polymerization rate constants (kobs) at different DBU loading. The plot of kobs versus the initial DBU concentration (Figure 2B) can also be linearly fitted, indicating that the polymerization is also firstorder dependent on the initiator concentration with a secondorder propagating rate constant (kp) of 5.05 ± 0.09 M−1 min−1. A close inspection of the ln([M]0:[M]) versus time plots revealed a slightly concave shape at the initial stage of the reactions, suggesting the initiation rate is either comparable to or slower than the rate of propagation. Thus, the kp obtained by the above kinetic analysis is somewhat approximate. To determine the relative initiation and propagation rate, a kinetic model that describes a living polymerization with slow initiation relative to propagation was further used to analyze the conversion versus reaction time plot for a selected polymerization ([M]0:[DBU]0 = 150:1, 50 °C, toluene-d8).52 The plot of the concentration of consumed monomers ([M]p) within the initial 25% conversion versus time exhibits good linearity (Figure 2C). The model fitting of the kinetic data yields an initiation rate constant (ki = 3.22 ± 1.29 M−1 min−1) that is nearly identical to the propagation rate constant (kp = 2.38 ± 0.03 M−1 min−1). To further validate these rate constants, a plot of conversion versus time up to 80% conversion was also fitted

reaching quantitative conversion revealed the absence of any free DBU initiators, suggesting that all DBUs ultimately became incorporated into the polymer chains through initiation. The solvent effect on the Mn and PDI control of the polymerization was investigated by conducting DBU-mediated polymerizations of Bu-NCA with various initial monomer to DBU ratios in a less polar solvent (toluene) and a more polar solvent (DMF) relative to THF. In toluene, the DBU-mediated polymerization of Bu-NCA exhibited a similar level of control over Mn and PDI as those conducted in THF (Table S1). By contrast, polymerization of Bu-NCA in DMF under identical conditions produced only low molecular weight PNBGs in the 0.5−4.3 kg mol−1 albeit with narrow molecular distribution (PDI = 1.04−1.11) (entries 5−9, Table S1). We attributed this to the competitive intramolecular backbiting by transamidation relative to chain propagation in DMF. The liberated DBU can reinitiate chain propagation, leading to high monomer conversion but low polymer molecular weight. This is supported by MALDI-TOF MS analysis of the resulting polymers, which consist of mainly cyclic PNBGs without DBU attached and minor cyclic PNBGs having one DBU moiety attached. Kinetic Study of DBU-Mediated Polymerization of BuNCA. We investigated the polymerization kinetics by conducting the DBU-mediated polymerization of Bu-NCA in toluene-d8 at 50 °C with a constant initial monomer concentration ([M]0 = 0.15 M) and varying initial monomer D

DOI: 10.1021/acs.macromol.5b02611 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

molecular weight range. For example, chain extension produces PNBG polymers whose molecular weight agrees reasonably well with the theoretical values based on efficient initiation by the PNBG macroinitiators at the low PNBG molecular range (DPn < 100, Table 2, entry 1). At the high MW range (DPn > 100), the MW of the resulting polymer is substantially lower than the theoretical value after the chain extension, suggesting the presence of competing side reactions. Characterization of Polymer Architecture by AFM and SANS. DBU-mediated polymerization of R-NCAs is expected to proceed through a cyclic zwitterionic propagating intermediate in the solvent with low dielectric constants (THF or toluene) similarly to the NHC-mediated polymerization.36 Therefore, the resulting PNBG polymers are hypothesized to have cyclic architectures. Experimental verification of the molecular architecture of the polymers proves to be nontrivial. The commonly used method is by measuring the intrinsic viscosity, hydrodynamic radius (Rh), or radius of gyration (Rg) of architecturally disparate polymers under θ conditions and comparing them to the linear counterpart having identical polymer molecular weight.19,27 Thus, the analysis conducted under non-θ conditions can only allow for qualitative assessment of polymer architecture. This approach can become misleading if the sample contains a mixture of different molecular architectures. It was previously demonstrated that the atomic force microscopy (AFM) can be used to reveal the backbone architecture (e.g., linear, cyclic, branched, etc.) of bottlebrush polymers.53−59 High grafting density of the polymeric side chains rigidifies the backbone conformations of bottlebrush polymers, making it possible to unambiguously identify the backbone architecture by AFM analysis. As a result, we synthesized a polymer bottlebrush consisting of polypeptoid backbone and poly(ethylene glycol) (PEG550) side chains by adapting a previously reported procedure.54 Specifically, a random copolymer consisting of poly(N-butylglycine) and poly(N-propargylglycine) [DBU-P(NBG41-r-NPgG177)] was first synthesized by the DBU-mediated copolymerization of Bu-NCA and Pg-NCA in THF (Scheme 2) and was subsequently conjugated with azido-terminated PEG550 (Mn = 550 kg mol−1) via copper-mediated alkyne−azido cycloaddition (CuAAC) reaction (Scheme 2). The choice of random copolypeptoid backbone is based on our previous observation that the PNBG-r-PNPgG copolypeptoids afforded higher PEG grafting efficiency using CuAAC chemistry than the poly(Npropargylglycine) homopolymers (PNPgG), as the latter has strong tendency to aggregate in solution.54 The 1H NMR spectrum of the cyclic bottlebrush polymers [DBU-P(NBG41-r-NPgG177)-g-(PEG550)177] revealed the appearance of triazole proton at 8.0 ppm and methylene protons of PEG at around 3.6 ppm (Figure S5). SEC analysis revealed that the bottlebrush polymer has a shorter elution time than the DBU-P(NBG41-r-NPgG177) precursor (Figure S6). These

using the kinetic model (Figure 2D), which gives the propagation rate constant of kp = 2.36 ± 0.07 M−1 min−1 on par with the kp obtained by fitting the [M]p versus time plot in the initial 25% conversion range (Figure 2B). As a result, the kinetic study suggests that the DBU-mediated polymerization of Bu-NCA undergoes the initiation event that is comparable in rate relative to propagation. The propagation rate is slightly faster than that of the previously studied ZROP of Bu-NCA using NHC initiators (kp = 1.53 ± 0.10 M−1 min−1).36 The difference in the propagation rate constant in the two systems can be attributed to the difference of counterion effect in the zwitterionic propagating intermediates (Scheme 1). DBU moieties on the cyclic zwitterionic propagating intermediate is less sterically demanding than the NHC moieties, thereby allowing for slightly faster addition of the monomer to the zwitterionic chain ends and a faster propagation rate in the former than the latter. Chain Extension Experiment. An important feature of controlled polymerization is the living characteristic of the propagating intermediates, allowing for chain extension with additional monomers. Two chain extension experiments were conducted using two PNBG macroinitiators with different chain length (DPn = 25 and 200) prepared by the DBUmediated polymerization of Bu-NCA. Additional Bu-NCA ([M]0:[PNBG]0 = 65:1 and 350:1) was introduced into the PNDG25 and PNDG200 macroinitiators, respectively. Both reactions (in 50 °C THF) were allowed to reach a complete conversion. SEC analysis of the PNBG polymers after the reaction revealed a single peak with low PDI (Table 2) at a Table 2. Chain Extension Polymerization from DBUTerminated PNBG Macroinitiatorsa entry no. 1

2

experiment before chain extension after chain extension before chain extension after chain extension

[M]0:[I]0

Mn(theor) (kg mol−1)

Mn(SEC)b (kg mol−1)

PDIb

25

3.0

3.0

1.10

90

10.2

9.2

1.02

200

22.7

17.2

1.05

550

62.3

37.4

1.10

a

All polymerization reactions were allowed to reach complete conversion. bThe measured Mn and PDI were obtained by the SECMALS-DRI method (DMF/0.1 M LiBr, 25 °C) with a measured dn/ dc = 0.0797 (9) mL g−1.

shorter elution time relative to the PNBG macroinitiator precursor (Figure S3), confirming the success of the chain extension and the living characteristic of the PNBG precursors. While PNBG produced by DBU-mediated polymerization of Bu-NCA can be further enchained, the control of the resulting polymer chain length appears to be dependent on the polymer Scheme 2

E

DOI: 10.1021/acs.macromol.5b02611 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 3. Toroid-shape bottlebrush polymer of DBU-P(NBG41-r-NPgG177)-g-(PEG550)177 viewed in representative AFM amplitude images. (A) 4 × 4 μm2; (B) 2.5 × 2.5 μm2.

combined results confirmed the successful synthesis of the bottlebrush polymers. Samples of the bottlebrush polymers exhibited toroid structures with a relatively uniform distribution across the flat mica substrate (Figure 3). The diameter and width of the rings measured 61 ± 11 nm and 23 ± 9 nm, respectively (Figure S7A,B). The dimensions measured with AFM are larger than the theoretically estimated diameter (d = 35 nm) and width (l = 9 nm) assuming a fully extended backbone and side chain structures. The discrepancy can be attributed to the effect of AFM tip−sample convolution.60 For comparison, we also synthesized the linear analogue of the bottlebrush polymer [P(NBG33-r-NPgG184)-g-(PEG550)184] consisting of a linear P(NBG33-r-NPgG184) backbone and PEG550 side chains by a reported procedure that involves the controlled polymerization of Bu-NCA and Pg-NCA using a primary amine initiator (BnNH2) and CuAAC chemistry.54 Samples of the linear architectural analogue were observed exclusively as rodlike structures with an average length and width of 208 ± 100 nm and 41 ± 12 nm, respectively (Figure S7C). The theoretically estimated length and width of the linear bottlebrush polymer should be 82 nm and 9 nm, assuming a fully stretched polymer backbone and side chains. This suggests that the rod structures are likely the aggregates of the linear bottlebrush polymers. To further verify the cyclic architecture of the bottlebrush polymers having the DBU-P(NBG41-r-NPgG177) backbone, small-angle neutron scattering (SANS) was also conducted on dilute solutions of the cyclic and linear bottlebrush polymer ensembles, respectively. The concentration of the solutions was kept in dilute regime (1 wt %) to ensure that the scattering data reflects the intramolecular conformation without interference from intermolecular interaction. To obtain the Kratky plots, the incoherent background, determined by fitting the high Q region of the scattering profile using a power function plus a constant, is subtracted from the original I(Q) before multiplying the square of momentum transfer (Q). The toroid-shape structure can be identified by the pronounced peak at Q = 0.06 Å−1 in the Kratky plot (red circle, Figure 4), which is attributed to the strong intramolecular correlation of the cyclic backbone structure.61 By contrast, the peak of the linear bottlebrush analogue in Kratky plot (black circle, Figure 4) was much less pronounced. Considering that the theoretically estimated

Figure 4. Kratky plots of cyclic bottlebrush DBU-P(NBG41-rNPgG177)-g-(PEG550)177 (red circle) and the linear analogue P(NBG33-r-NPgG184)-g-(PEG550)184 (black circle).

length of the PEG side chains is 4.5 nm, the small bump in the Kratky plot of the linear analogue may be attributed to the weak intramolecular correlation of the PEG side chains. The SANS results are consistent with the cyclic architecture of polypeptoid backbone synthesized by the DBU-mediated ZROP of R-NCAs. Detailed SANS analyses of the cyclic and linear bottlebrush copolymer structures are currently in progress. Investigation of the Initiation Event. To gain insights into the initiation step of the DBU-mediated polymerization of R-NCAs, DBU and Bu-NCA in a 1:1 stoichiometric ratio were allowed to react in THF at 50 °C for 2 h. 1H NMR analysis revealed the complete disappearance of the reactants. The product was isolated as a brownish solid by precipitation in hexane. The resulting product was analyzed by ESI-MS spectroscopy in positive ionization mode. Two molecular species whose mass are consistent with cyclic trimeric or tetrameric N-butylglycine N-carboxyanhydride species are observed in ESI-MS spectrum (Figure S8). This agrees with kinetic analysis, which reveals an initiation step that is comparable in rate to propagation. This makes isolation of initiating species challenging. Apart from these two species, a third molecular ion consistent with the protonated DBU is also present in the ESI-MS spectrum. The species observed in ESI-MS analysis differ from what was F

DOI: 10.1021/acs.macromol.5b02611 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Scheme 3

intermediate species 2. The carboxylic−carbamic anhydride linkages in 2 is thermally labile and readily undergoes intramolecular decarboxylation to yield 3, which has been observed spectroscopically in the early studies of small model compounds.66,67 Species 3 can undergo decarboxylation to form 4, which is observed in the MALDI-TOF MS analysis. We propose that the decarboxylation is reversible, allowing 4 to regain CO2 to form 3, from which monomer addition and chain propagation ensue. This is consistent with the living polymerization characteristics of the DBU-mediated polymerization of Bu-NCA. The formation of species 5 presumably from macrocyclizaion of 3 is supported by the ESI MS analysis. Its formation is not kinetically competitive relative to monomer addition, thus giving rise to the controlled polymerization behavior. Furthermore, the methylene proton on the cationic DBU end-group that is adjacent to the positively charged carbon is sufficiently acidic to be deprotonated by anionic alkoxide, as shown in a previous study of DBU-mediated zwitterionic polymerization of lactide, resulting in the formation of linear polylactide impurities.50 In this study, we do not observe the deprotonation of the methylene protons in the DBU end-group based on the 1H NMR analysis of the 1:1 reaction of Bu-NCA and DBU. This is presumably due to the reduced basicity of the anionic carbamate chain end relative to the alkoxide species.

observed in the MALDI-TOF MS analysis of a low molecular PNBG (DPn = 25) obtained from DBU-mediated polymerization of Bu-NCA where the major molecular ions are due to the cyclic PNBGs bearing one DBU moiety. The discrepancy may arise from the difference in MS conditions: electrospray ionization can cause elimination of the initiating moieties (i.e., DBU in this study) from the cyclic zwitterionic polymeric species as previously demonstrated in the detailed MS studies of cyclic PNBG polymers obtained from NHC-mediated ZROP of Bu-NCAs.51 FT-IR spectroscopic analysis of the species obtained from the 1:1 molar ratio reaction of DBU and Bu-NCA further confirms the formation of zwitterionic species in the initiation event (Figure S9). The peak at 1572 cm−1 is due to the carbonyl stretch of the carbamate group at the chain end,36,62,63 which is consistent with the molecular ions observed by the ESI-MS analysis of the reaction mixture. The peak at 1684 cm−1 is due to the amide carbonyl group of the PNBG backbone. The amide peak that is immediate adjacent to the cationic DBU moiety is blue-shifted to at 1755 cm−1 presumably due to the inductive effect of the cationic DBU moiety. The peak at 1646 and 1613 cm−1 are due to CN stretching mode of the positively charged DBU moiety and the unreacted DBU, respectively.64,65 As a comparison, the FTIR spectrum of the species obtained from the reaction of benzyl amine (BnNH2) and Bu-NCA in a 1:1 molar ratio exhibits only one carbonyl peak (1680 cm−1) which is due to the amide carbonyl of the linear PBNG backbone, clearly indicating a different initiation and propagation mechanism from that of the DBU-mediated polymerization of Bu-NCA. It is proposed that DBU initiates the polymerization by nucleophilic ring-opening addition of Bu-NCA at the 5carbonyl position to form the zwitterionic initiating species 1 (Scheme 3). The chain propagation ensues by Bu-NCA monomer addition at the carboxylate end of 1 to form the



CONCLUSIONS

DBU-mediated ZROPs of Bu-NCAs exhibit characteristics of a controlled polymerization, producing poly(N-butylglycine) with controlled polymer molecular weight (3.5−32.4 kg mol−1) and narrow molecular weight distribution (PDI < 1.2). The growing chain ends can be successfully extended with additional monomers, making it useful for the block copolypeptoid synthesis. Kinetic studies revealed that the initiation rate is comparable to the propagation rate in the G

DOI: 10.1021/acs.macromol.5b02611 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

(6) Kirshenbaum, K.; Barron, A. E.; Goldsmith, R. A.; Armand, P.; Bradley, E. K.; Truong, K. T.; Dill, K. A.; Cohen, F. E.; Zuckermann, R. N. Proc. Natl. Acad. Sci. U. S. A. 1998, 95, 4303−4308. (7) Lee, B.-C.; Zuckermann, R. N.; Dill, K. A. J. Am. Chem. Soc. 2005, 127, 10999−11009. (8) Guo, L.; Li, J.; Brown, Z.; Ghale, K.; Zhang, D. Biopolymers 2011, 96, 596−603. (9) Rosales, A. M.; Murnen, H. K.; Kline, S. R.; Zuckermann, R. N.; Segalman, R. A. Soft Matter 2012, 8, 3673−3680. (10) Baldauf, C.; Gunther, R.; Hofmann, H. J. Phys. Biol. 2006, 3, 1− 9. (11) Armand, P.; Kirshenbaum, K.; Goldsmith, R. A.; Farr-Jones, S.; Barron, A. E.; Truong, K. T. V.; Dill, K. A.; Mierke, D. F.; Cohen, F. E.; Zuckermann, R. N.; Bradley, E. K. Proc. Natl. Acad. Sci. 1998, 95, 4309−4314. (12) Mannige, R. V.; Haxton, T. K.; Proulx, C.; Robertson, E. J.; Battigelli, A.; Butterfoss, G. L.; Zuckermann, R. N.; Whitelam, S. Nature 2015, 526, 415−420. (13) Guo, L.; Li, J.; Brown, Z.; Ghale, K.; Zhang, D. Pept. Sci. 2011, 96, 596−603. (14) Tanisaka, H.; Kizaka-Kondoh, S.; Makino, A.; Tanaka, S.; Hiraoka, M.; Kimura, S. Bioconjugate Chem. 2008, 19, 109−117. (15) Makino, A.; Kizaka-Kondoh, S.; Yamahara, R.; Hara, I.; Kanzaki, T.; Ozeki, E.; Hiraoka, M.; Kimura, S. Biomaterials 2009, 30, 5156− 5160. (16) Lahasky, S. H.; Hu, X.; Zhang, D. ACS Macro Lett. 2012, 1, 580−584. (17) Ulbricht, J.; Jordan, R.; Luxenhofer, R. Biomaterials 2014, 35, 4848−4861. (18) Clarson, S.; Semlyen, J. Polymer 1986, 27, 1633−1636. (19) Bielawski, C. W.; Benitez, D.; Grubbs, R. H. Science 2002, 297, 2041−2044. (20) Kricheldorf, H. R. J. Polym. Sci., Part A: Polym. Chem. 2010, 48, 251−284. (21) Shin, E. J.; Jeong, W.; Brown, H. A.; Koo, B. J.; Hedrick, J. L.; Waymouth, R. M. Macromolecules 2011, 44, 2773−2779. (22) Qiu, X.-P.; Tanaka, F.; Winnik, F. M. Macromolecules 2007, 40, 7069−7071. (23) Ishizu, K.; Ichimura, A. Polymer 1998, 39, 6555−6558. (24) Xia, Y.; Boydston, A. J.; Yao, Y.; Kornfield, J. A.; Gorodetskaya, I. A.; Spiess, H. W.; Grubbs, R. H. J. Am. Chem. Soc. 2009, 131, 2670− 2677. (25) Laurent, B. A.; Grayson, S. M. J. Am. Chem. Soc. 2006, 128, 4238−4239. (26) Yamamoto, T.; Tezuka, Y. Polym. Chem. 2011, 2, 1930−1941. (27) Culkin, D. A.; Jeong, W.; Csihony, S.; Gomez, E. D.; Balsara, N. P.; Hedrick, J. L.; Waymouth, R. M. Angew. Chem., Int. Ed. 2007, 46, 2627−2630. (28) Jeong, W.; Shin, E. J.; Culkin, D. A.; Hedrick, J. L.; Waymouth, R. M. J. Am. Chem. Soc. 2009, 131, 4884−4891. (29) Brown, H. A.; Chang, Y. A.; Waymouth, R. M. J. Am. Chem. Soc. 2013, 135, 18738−18741. (30) Brown, H. A.; Xiong, S.; Medvedev, G. A.; Chang, Y. A.; AbuOmar, M. M.; Caruthers, J. M.; Waymouth, R. M. Macromolecules 2014, 47, 2955−2963. (31) Brown, H. A.; Waymouth, R. M. Acc. Chem. Res. 2013, 46, 2585−2596. (32) Acharya, A. K.; Chang, Y. A.; Jones, G. O.; Rice, J. E.; Hedrick, J. L.; Horn, H. W.; Waymouth, R. M. J. Phys. Chem. B 2014, 118, 6553− 6560. (33) Zhang, X.; Waymouth, R. M. ACS Macro Lett. 2014, 3, 1024− 1028. (34) Asenjo-Sanz, I.; Veloso, A.; Miranda, J. I.; Pomposo, J. A.; Barroso-Bujans, F. Polym. Chem. 2014, 5, 6905−6908. (35) Shin, E. J.; Brown, H. A.; Gonzalez, S.; Jeong, W.; Hedrick, J. L.; Waymouth, R. M. Angew. Chem., Int. Ed. 2011, 50, 6388−6391. (36) Guo, L.; Lahasky, S. H.; Ghale, K.; Zhang, D. J. Am. Chem. Soc. 2012, 134, 9163−9171.

DBU-mediated ZROP of Bu-NCA, which is slightly faster than previously reported NHC-mediated ZROP. This is presumably due to the steric difference of the DBU versus the NHC moieties in the respective cyclic zwitterionic propagating intermediate. AFM and SANS studies confirm the cyclic architectures of the polypeptoids synthesized by the DBUmediated ZROPs of R-NCAs. The enhanced robustness and availability of DBU relative to NHCs and the excellent control over the polymerization of R-NCA make the reported method an attractive route toward well-defined cyclic polypeptoids with tunable ring sizes.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.5b02611. 1 H NMR and MALDI-TOF MS spectra of a low MW PNBG polymer obtained by DBU-mediated polymerization of Bu-NCA in THF; a table that summarizes the MW and PDI of PNBGs obtained by DBU-mediated polymerizations of Bu-NCAs in toluene or DMF; SEC chromatograms of PNBGs obtained by DBU-mediated polymerization of Bu-NCA and chain-extension experiment; 1H NMR spectrum of DBU-P(NBG41-r-NPgG177) obtained by DBU-mediated polymerization of Bu-NCA and Pg-NCA; 1H NMR spectrum and SEC chromatogram of the cyclic bottlebrush polymer DBU-P(NBG41-rNPgG177)-g-(PEG550)177 and the DBU-P(NBG41-rNPgG177) precursor; representative AFM images of the cyclic and linear bottlebrush polymers DBU-P(NBG41-rNPgG177)-g-(PEG550)177 and P(NBG30-r-NPgG180)-g(PEG550)180; ESI-MS and FT-IR spectra of the product obtained from the reaction of Bu-NCA and DBU (or BnNH2) in a 1:1 initial molar ratio (PDF)



AUTHOR INFORMATION

Corresponding Author

*(D.Z.) E-mail [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work is supported by the National Science Foundation (CHE 0955820) and LSU. The SANS studies are supported by the U.S. Department of Energy under EPSCoR Grant DESC0012432 with additional support from the Louisiana Board of Regents. The Research at Oak Ridge National Laboratory’s High Flux Isotope Reactor and Spallation Neutron Source was sponsored by the Scientific User Facilities Division, Office of Basic Energy Sciences, U.S. Department of Energy.



REFERENCES

(1) Miller, S. M.; Simon, R. J.; Ng, S.; Zuckermann, R. N.; Kerr, J. M.; Moos, W. H. Bioorg. Med. Chem. Lett. 1994, 4, 2657−2662. (2) Rosales, A. M.; Murnen, H. K.; Zuckermann, R. N.; Segalman, R. A. Macromolecules 2010, 43, 5627−5636. (3) Lee, C.-U.; Smart, T. P.; Guo, L.; Epps, T. H., III; Zhang, D. Macromolecules 2011, 44, 9574−9585. (4) Lee, C.-U.; Li, A.; Ghale, K.; Zhang, D. Macromolecules 2013, 46, 8213−8223. (5) Guo, L.; Zhang, D. J. Am. Chem. Soc. 2009, 131, 18072−18074. H

DOI: 10.1021/acs.macromol.5b02611 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules (37) Kricheldorf, H. R.; Bösinger, K. Makromol. Chem. 1976, 177, 1243−1258. (38) Kricheldorf, H. R.; Von Lossow, C.; Schwarz, G. J. Polym. Sci., Part A: Polym. Chem. 2006, 44, 4680−4695. (39) Kricheldorf, H. R.; von Lossow, C.; Schwarz, G. Macromolecules 2005, 38, 5513−5518. (40) Kricheldorf, H. R.; Lossow, C. V.; Lomadze, N.; Schwarz, G. J. Polym. Sci., Part A: Polym. Chem. 2008, 46, 4012−4020. (41) Ono, N.; Yamada, T.; Saito, T.; Tanaka, K.; Kaji, A. Bull. Chem. Soc. Jpn 1978, 51, 2401−2404. (42) Lohmeijer, B. G.; Pratt, R. C.; Leibfarth, F.; Logan, J. W.; Long, D. A.; Dove, A. P.; Nederberg, F.; Choi, J.; Wade, C.; Waymouth, R. M. Macromolecules 2006, 39, 8574−8583. (43) Coady, D. J.; Fukushima, K.; Horn, H. W.; Rice, J. E.; Hedrick, J. L. Chem. Commun. 2011, 47, 3105−3107. (44) Iwasaki, Y.; Yamaguchi, E. Macromolecules 2010, 43, 2664− 2666. (45) Todd, R.; Templaar, S.; Lo Re, G.; Spinella, S.; McCallum, S. A.; Gross, R. A.; Raquez, J.-M.; Dubois, P. ACS Macro Lett. 2015, 4, 408− 411. (46) De Rycke, N.; Couty, F.; David, O. R. Chem. - Eur. J. 2011, 17, 12852−12871. (47) Baidya, M.; Mayr, H. Chem. Commun. 2008, 1792−1794. (48) Carafa, M.; Mesto, E.; Quaranta, E. Eur. J. Org. Chem. 2011, 2011, 2458−2465. (49) Shieh, W.-C.; Dell, S.; Repic, O. J. Org. Chem. 2002, 67, 2188− 2191. (50) Brown, H. A.; De Crisci, A. G.; Hedrick, J. L.; Waymouth, R. M. ACS Macro Lett. 2012, 1, 1113−1115. (51) Li, X.; Guo, L.; Casiano-Maldonado, M.; Zhang, D.; Wesdemiotis, C. Macromolecules 2011, 44, 4555−4564. (52) Wang, L.; Bochmann, M.; Cannon, R. D.; Carpentier, J.-F.; Roisnel, T.; Sarazin, Y. Eur. J. Inorg. Chem. 2013, 2013, 5896−5905. (53) Tang, H.; Li, Y.; Lahasky, S. H.; Sheiko, S. S.; Zhang, D. Macromolecules 2011, 44, 1491−1499. (54) Lahasky, S. H.; Serem, W. K.; Guo, L.; Garno, J. C.; Zhang, D. Macromolecules 2011, 44, 9063−9074. (55) Xia, Y.; Boydston, A. J.; Grubbs, R. H. Angew. Chem., Int. Ed. 2011, 50, 5882−5885. (56) Boydston, A. J.; Holcombe, T. W.; Unruh, D. A.; Fréchet, J. M.; Grubbs, R. H. J. Am. Chem. Soc. 2009, 131, 5388−5389. (57) Schappacher, M.; Deffieux, A. J. Am. Chem. Soc. 2008, 130, 14684−14689. (58) Schappacher, M.; Deffieux, A. Angew. Chem. 2009, 121, 6044− 6047. (59) Schappacher, M.; Deffieux, A. Science 2008, 319, 1512−1515. (60) Villarrubia, J. J. Res. Natl. Inst. Stand. Technol. 1997, 102, 425− 454. (61) Gagliardi, S.; Arrighi, V.; Ferguson, R.; Dagger, A. C.; Semlyen, J. A.; Higgins, J. S. J. Chem. Phys. 2005, 122, 064904. (62) Choi, B. G.; Kim, G. H.; Yi, K. B.; Kim, J.-N.; Hong, W. H. Korean J. Chem. Eng. 2012, 29, 478−482. (63) Qi, G.; Wang, Y.; Estevez, L.; Duan, X.; Anako, N.; Park, A.-H. A.; Li, W.; Jones, C. W.; Giannelis, E. P. Energy Environ. Sci. 2011, 4, 444−452. (64) Galezowski, W.; Jarczewski, A.; Stanczyk, M.; Brzezinski, B.; Bartl, F.; Zundel, G. J. Chem. Soc., Faraday Trans. 1997, 93, 2515− 2518. (65) Ferri, D.; Bürgi, T.; Baiker, A. J. Chem. Soc., Perkin Trans. 2 2002, 437−441. (66) Rawlinson, D. J.; Humke, B. M. Tetrahedron Lett. 1972, 13, 4395−4398. (67) Dell’Amico, D. B.; Calderazzo, F.; Giurlani, U. J. Chem. Soc., Chem. Commun. 1986, 1000−1001.

I

DOI: 10.1021/acs.macromol.5b02611 Macromolecules XXXX, XXX, XXX−XXX